Photoelectron Spectroscopy of the Hexafluorobenzene Cluster Anions

Subscriber access provided by University of Winnipeg Library

Article

Photoelectron Spectroscopy of the Hexafluorobenzene Cluster Anions: (C6F6)n- (n = 1 – 5) and I-(C6F6) Joshua P. Rogers, Cate Sara Anstöter, James N Bull, Basile F. E. Curchod, and Jan R. R. Verlet J. Phys. Chem. A, Just Accepted Manuscript • DOI: 10.1021/acs.jpca.8b11627 • Publication Date (Web): 29 Jan 2019 Downloaded from http://pubs.acs.org on January 30, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Photoelectron Spectroscopy of the Hexafluorobenzene Cluster Anions: (C6F6)n− (n = 1 – 5) and I−(C6F6) Joshua P. Rogers, Cate S. Anstöter, James N. Bull, Basile F. E. Curchod and Jan R. R. Verlet*

Department of Chemistry, Durham University, Durham DH1 3LE

ABSTRACT Frequency-resolved (2D) photoelectron (PE) spectra of the anionic clusters (C6F6)n−, for n = 1 – 5, and time-resolved PE spectra of I−C6F6 are presented using a newly built instrument and supported by electronic structure calculations. From the 2D PE spectra, the vertical detachment energy (VDE) of C6F6− was measured to be 1.60 ± 0.01 eV and the adiabatic detachment energy (ADE) ≤ 0.70 eV. The PE spectra also contain fingerprints of resonance dynamics over certain photon energy ranges, in agreement with the calculations. An action spectrum over the lowest resonance is also presented. The 2D spectra of (C6F6)n− show that the cluster can be described as C6F6−(C6F6)n−1. The VDE increases linearly (200 ± 20 meV n−1) due to the stabilising influence on the anion of the solvating C6 F6 molecules. For I−C6F6, action spectra of the absorption just below both detachment channels are presented. Timeresolved PE spectra of I−C6F6 excited at 3.10 eV and probed at 1.55 eV reveal a short-lived non-valence state of C6F6− that coherently evolves into the valence ground state of the anion and induces vibrational motion along a specific buckling coordinate. Electronic structure calculations along the displacement of this mode show that at the extreme buckling angle, the probe can access an excited state of the anion that is bound at that geometry, but adiabatically unbound. Hence, slow electrons are emitted and show dynamics that probe predominantly the outer-turning point of the motion. A PE spectrum taken at t = 0 contains vibrational structure, assigned to a specific Raman and/or IR active mode of C6F6.

e-mail: [email protected]

1 ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 37

INTRODUCTION Hexafluorobenzene, C6F6, has been extensively studied in electron attachment experiments as a text-book case in which the additional electron captured in the valence orbital of the anion leads to a dramatic Jahn-Teller distortion of the planar D6h geometry associated with the neutral1-7. More recently, it has been theoretically suggested that the C6F6 anion, C6F6−, also possesses a non-valence state for which the electron binding is predominantly driven by correlation forces8. This diffuse orbital associated with the neutral geometry of C6F6 provides a door-way state to the formation of valence-bound C6F6− and the dynamics of this attachment process have recently been probed in real-time9. Beyond the interaction of a single electron with a single C6F6 molecule, C6F6 also has some remarkable bulk properties. The electron mobility in liquid C6F6 is an order of magnitude larger than that expected from the motion of molecular anions through the liquid,2 suggesting that the excess electron passes from one molecule to its neighbour without causing a change in the geometry of either molecule from a neutral planar configuration. Additionally, C6F6 is an exceedingly efficient scavenger of excess electrons in the non-attaching molecular liquid tetramethylsilane2, again suggesting an efficient electron accepting pathway. In a recent study, we have shown evidence for non-valence states in anionic clusters of C6F6, where a low energy outgoing electron produced by photodetachment from (C6F6)n− (where n = 2 – 5) was observed to be recaptured, leading to an indirect electron loss channel10. Finally, twophoton photoemission and scanning tunnelling microscopy experiments of C6F6 on Cu(111) surfaces point to the formation of diffuse anion states during the electron-capture process3,11. Given the interest in electron attachment to C6F6, there have been a number of studies3,12-14 on C6F6−, including two PE spectroscopy studies15,16. Kaya and coworkers measured the PE spectra of (C6F6)n− with n = 1 – 8 at a photon energy, hv = 3.49 eV, as well as that of AuC6F6−. Later, Bowen and coworkers studied C6F6−(H2O)n with n = 0 – 2 at hv =


Page 3 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


2.54 eV. Both studies found the PE spectra of C6F6− to be very broad and with a vertical detachment energy of 1.55 eV. While the two experiments had overall agreement, the PE spectra of Bowen and coworkers showed additional vibrational structure on the broad PE profile. In the present study, we extend these previous works and probe the PE at a range of different hv as well as probing to higher hv, where new detachment channels become accsessible. Performing PE spectroscopy over a range of hv can provide complementary information regarding the electron attachment process by virtue of its half-reaction equivalent17-19. Specifically, it is complementary to 2D electron energy loss spectroscopy and can access those resonances formed in electron impact, with the important caveat that the selection rules differ and the initial molecular geometries differ17. For C6F6−, there are large changes in geometries, causing this complementarity to break down. Nevertheless, as will be discussed below, the 2D PE spectra provide information about resonances accessible from the anion ground state. Our PE spectra of C6F6− are broadly in agreement with the previous studies, although we do not observe the vibrational structure seen by Bowen and coworkers. Moreover, we collect the PE angular distributions and these are also consistent with a JahnTeller distorted geometry. We observed the lowest triplet state of C6F6 and show that it has bent geometry, similar to that of C6F6−. Additionally, there is evidence for resonances in C6F6−, which we assign with the aid of electronic structure calculations. Finally, we also show results from I−C6F6 in which we excite a charge-transfer band that accesses the nonvalence state9. The PE spectrum of the non-valence state is tentatively assigned and we analyse the time-resolved conversion from non-valence to valence with the aid of the electronic structure of C6F6− determined from the 2D PE spectra and computational results.



Page 4 of 37

EXPERIMENTAL AND COMPUTATIONAL DETAILS Anion photoelectron spectrometer

Figure 1: Overview of anion photoelectron imaging instrument. A schematic of the top-view of the instrument is shown in (a) where A is the ion source, B the time-of-flight plates of the mass-spectrometer, and C the velocity map imaging PE spectrometer, which is also shown as a side-view. (b) shows the electron beam profile in the expansion. (c) shows the PE spectrum of O2− derived from the PE image with the left half being the raw (2D crushed) PE image and the right half the reconstructed central slice through the 3D PE distribution.

We have recently developed a new anion PE spectrometer, which we will detail here for the first time. An overview of the instrument is shown in Figure 1(a). It consists of molecular beam anion source followed by a time-of-flight (TOF) mass spectrometer coupled to a velocity-map imaging (VMI) PE spectrometer20. A supersonic expansion is formed by expanding seeded analyte molecules in a carrier gas using a high temperature Even-Lavie pulsed valve21. The expansion is crossed with a focussed electron beam (1 mA beam current 4 ACS Paragon Plus Environment

Page 5 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


with a full width at half maximum of 0.6 mm and 300 eV kinetic energy – see Figure 1(b)), generated by a home-built electron gun. The location of ionisation within the supersonic expansion is adjustable by translating the valve relative to the fixed electron beam axis. The supersonic expansion passes through a skimmer and a series of tubular collimating electrodes. After travelling 0.38 m, the gas cloud enters a Wiley-McLaren TOF assembly22, which is rapidly switched to a high negative voltage ( 5 eV. These are presented more clearly in Figures 4(b) and (c). Both PE features show vibrational structure. According to our electronic structure calculations (Figure 6), these features correspond to the two lowest triplet states of the C6F6, T1 and T2. From the PE spectra, we find that the adiabatic energy of the T1 and T2 states relative to the anion ground state to be 3.73±0.10 and 4.70±0.10 eV, respectively. This compares reasonably well with their calculated energies at 4.26 and 4.99 eV, respectively. The fact that vibrational structure is observed, especially for the T2 excited state, points to a comparatively small geometric difference between the two lowest triplet states and the anion buckled geometry (shown in Figure 5(b)). Indeed, our calculations show that the lowest energy structures for T1 and T2 have buckled geometries. Note that the S1 state of neutral C6F6 is calculated to lie at 5.72 eV above the anion so that it is not seen in our experiment.



Page 14 of 37

Figure 7: (a) PE image of C6F6– taken at hv = 2.2 eV showing direct and indirect (boxed in the centre of image) detachment channels. Vertical double arrow indicates the polarisation axis of the light. (b) Action spectrum acquired by taking the ratio of indirect to direct detachment channels as a function of hv. In addition to the direct detachment channels (S0 + e−, T1 + e−, and T2 + e−), there is also evidence in the 2D-PE spectra that excited states (resonances) of C6F6− are excited. This can be seen in around hv = 3.0 eV and for hv < 2.4 eV. In the latter range, very low energy electrons are emitted, which can be appreciated from the raw PE image taken at hv = 2.2 eV shown in Figure 7(a). As the direct photodetachment channel is strongly anisotropic and predominantly of p-wave character, very few electrons should be emitted at threshold because of the threshold behaviour of photodetachment36. Yet, the PE image in Figure 7(a) shows some isotropic emission near threshold suggesting that there is a detachment process accessible via an indirect mechanism. We have crudely analysed the absorption (action) 14 ACS Paragon Plus Environment

Page 15 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


spectrum associated with the appearance of these low energy electrons by integrating the central area of the image (indicated in Figure 7(a) by a box) as a fraction of the total PE counts in the image. The results of this analysis are shown in Figure 7(b) over the range 1.9 < hv < 2.5 eV. The rising background towards lower hv is an experimental artefact that comes about from the fact that the PE signal in the direct detachment peak overlaps more with the boxed region of the image in Figure 7(a) at lower hv. Figure 7(b) shows a broad structureless peak centred at hv ~ 2.2 eV. The relative contribution of the indirect feature is in fact quite small, amounting to only ~ 1% of the direct PE signal. According to our calculations, the D1 excited state lies at 2.36 eV above the D0 (in the anion geometry) and could be a candidate for this resonance. While this transition has negligible oscillator strength at the optimised anion geometry on symmetry grounds (and was therefore not included in Figure 6), our calculations also show that it gains significant oscillator strength with small geometric perturbations of the lowest frequency modes. Therefore, this transition may be observed and the intensity seen in Figure 7(b) is consistent with the calculated small oscillator strength. Additional evidence for resonance dynamics can be seen between 2.7 < hv < 3.9 eV, where PE signal is observed at lower eKE than that of the direct S0 + e− channel. A representative PE spectrum at hv = 2.9 eV is shown in Figure 8. According to our calculations (Figure 6), there are several candidate resonances; there are three optically bright doublet states between 2.62 and 3.25 eV. The appearance of electrons at lower eKE most likely arises from nuclear motion on the potential energy surface of the resonance that leads to changes in the Franck-Condon overlap with the neutral ground state. This has been seen in several PE spectroscopic studies37-42. We also note that the direct detachment peak appears slightly shifted towards lower eKE, which again points to the presence of resonances43. We cannot identify which of the three bright resonances contribute to the spectral changes, but it



Page 16 of 37

is clear that above-threshold dynamics are occurring that lead to electron emission at lower eKE. The location of resonances of C6F6− have been studied using electron impact spectroscopy7. However, these are of limited use as a comparison to the current study because of the large structural difference between the initial geometries of the anion and neutral species.

Figure 8: PE spectrum of C6F6– taken at hv = 2.9 eV. Signal at eKE < 0.5 eV arises from indirect detachment following excitation to resonances.

Finally, we briefly comment on the apparent change in slope of the direct detachment channel leading to the S0 state around hv ~ 3 eV. Firstly, because of the resonances around 3 eV, the shape of the “direct” detachment also includes contributions from some indirect electron emission, which pushes the apparent VDE to lower eKE. Secondly, because of the width of the direct peak is so large, as hv becomes smaller, part of the peak is below threshold, which (because of the threshold behaviour) has the appearance of pushing the VDE to higher energy. Finally, it is also worth noting that the output of the OPO reaches a minimum at 3.10 eV so that there is increased noise at this hv.


Page 17 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


C6F6 anion clusters, (C6F6)n− Figure 9 shows the 2D-PE spectra for (C6F6)2−. Overall, these data appear similar to the 2D-PE spectra in Figure 3. Specifically, the same direct detachment channels (S0 + e−, T1 + e−, and T2 + e−) are present, although the vibrational structure in the triplet neutral states is less apparent. Additionally, the onset of the direct detachment channels has increased by ~200 meV compared to the unclustered C6F6− ion. This increase in VDE (and ADE) predominantly arises from the cohesion energy of the cluster. In addition to the direct detachment channels, there is again evidence of resonance dynamics occurring that leads to electron emission at lower eKE than expected for the direct S0 + e− channel. The range over which such emission is observed is broader in the dimer than in the monomer PE spectrum; weak signal occurs at most measured hv, with particularly clear evidence for resonance dynamics in the regions of hv ~ 2.8, 3.7, and 4.3 eV. We have not performed detailed calculations of the excited state (C6F6)2− and offer no assignment of the electronic states facilitating these dynamics.

Figure 9: 2D PE spectra of (C6F6)2− in the range 1.8 ≤ hv ≤ 5.8 eV at intervals of 0.1 eV. Each PE spectrum has been normalised to its total integrated signal.



Page 18 of 37

For hv > 4.2 eV, there appears to be enhanced signal at very low eKE. Such electrons are indicative of a delayed emission process in which the emission is becoming more statistical in nature, again pointing to dynamics involving resonances17. However, this delayed emission channel coincides with the opening of the T1 + e− and T2 + e− detachment channels. We have previously considered the origin of this signal and assigned it to an indirect mechanism in which the outgoing PE is captured by the non-valence anion state localised on the solvent molecule10. A similar onset of the appearance of low-eKE electrons at the opening of the T1 + e− and T2 + e− detachment channels was also observed for large clusters (with up to 5 molecules), suggesting a common mechanism in all clusters, which would be consistent with a picture in which the low-eKE outgoing electron is recaptured by a solvent molecule. Figure 10 shows a series of PE spectra for (C6F6)n− taken at hv = 3.80 eV, which clearly show that an incremental increase in cluster size leads to an increase in the VDE. In Figure 11, the measured increase in VDE is plotted as a function of cluster size, n. This shows a roughly linear trend with a gradient of 200 ± 20 meV n−1, indicating that photoemission occurs from C6F6− solvated in a C6F6 cluster: C6F6−(C6F6)n−1. This is consistent with the observation that for all (C6 F6)n−, the direct S0 + e− photodetachment is highly anisotropic peaking along the polarisation axis. We conclude therefore that the emitted PE originates from the SOMO of C6F6− (Figure 5), which, although energetically stabilised, is not substantially perturbed by the surrounding solvent cluster.


Page 19 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 10: PE spectra of (C6F6)n− (n = 1-5) taken at hv = 3.80 eV

The observed increase in VDE as a function of n is in agreement with that of Nakajima et al.16 over the same range (up to n = 5) as shown in Figure 11. Our data and that of Nakajima et al. suggests that the increase in VDE up to n = 5 is essentially linear. Beyond this range (up to n = 8), Nakajima et al. showed the incremental increase becomes smaller (see Figure 11). It is also interesting to note that the FWHM of the direct detachment peaks in Figure 10 appears to grow with increasing n. Although this trend is not seen in the measurements made by Nakajima et al., a direct comparison may not be appropriate because the two experiments were not performed at the same excitation energy and resonances may affect the apparent width of these features.



Page 20 of 37

Figure 11: Measured vertical detachment energies (VDE) for (C6F6)n− for n = 1-5 from the present study (black full circles) with a linear fit (sold line), and with n = 1-8 from Nakajima et al.16 (red open circles).

I−C6F6 cluster: Photoelectron and photodetachment spectroscopy The 2D-PE spectrum of I−(C6F6) has been presented previously9. In brief, this data showed that the charge is localised on the iodide and direct detachment essentially yields the PE spectrum of I−, shifted to higher eBE by the cluster binding energy with C6 F6. Hence, the I−(C6F6) complex can be viewed as I− solvated by a neutral (planar) C6F6 molecule. Below the adiabatic detachment energy, PE signal was observed and assigned to detachment from I−. Such signal can only arise from the excitation of the I−(C6F6) complex to some excited state that then dissociates to yield I− and a second photon within the ~5 ns pulse detaches an electron leading to the observed PE spectrum of I−. The excited state is a charge-transfer state, which injects the electron onto the planar C6F6. In principle, an action spectrum of this absorption band could be measured by monitoring the I− yield as a function of hv. However, we do not have the ability to measure secondary mass spectra in our current experiment. Instead, we have monitored the total electron yield as a function of hv. The electron yield following excitation of I−(C6F6) below the ADE is expected to depend quadratically on the


Page 21 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


laser intensity because of the two-photon (sequential) process. In Figure 12, the photodetachment action spectrum (square-root of total electron yield) is shown in the range 3.1 < hv < 3.5 eV. This spectrum clearly shows a peak around hv ~ 3.3 eV, below the ADE of the direct detachment channel (ADE = 3.48 eV) 9. We note, however, that the 355 nm pumped OPO is unstable around hv = 3.1 eV such that the rising edge of this was difficult to acquire confidently. Any apparent structure in the band is mostly noise. In order to confirm the measurement of an action spectrum below the onset of the I[2P3/2]∙C6F6 + e− channel, we have also acquired a low resolution action spectrum just below the onset of the second I[2P1/2]∙C6F6 + e− channel. Substantial low eKE PE signal was observed in the 2D-PE spectra in this range indicating an indirect electron loss mechanism, which in turn implies the presence a resonant excited state9. In order to determine the spectral shape of its absorption band, we performed the same analysis on these spectra as in Figure 7, comparing the yield of low eKE electrons to the PE signal arising from the direct detachment process. This ratio is represented by the red circles in Figure 12. The overall peak shape of the two action spectra match but are offset by 0.94 eV, in agreement with the spinorbit splitting in iodine. Although neither method is ideally suited to record an action spectrum, there is adequate qualitative agreement over the shape and location of the absorption bands. The peak of the absorption is ~0.2 eV below the opening of the respective detachment channels, suggesting that the excited state is bound by this amount. However, we also note that the peaks are broad and have a width of ~0.2 eV (full width at half maximum). It is not immediately clear why the absorption band should be this broad. Nevertheless, similar broad absorption bands have been seen in the action spectra of iodide with pyrrole44. In our timeresolved experiments (described below), the red-edge of the absorption band was excited with hv = 3.10 eV.



Page 22 of 37

Figure 12: Action (absorption) spectra of I−(C6F6) near the I[2P3/2]∙C6F6 + e− channel (solid black line) and the I[2P1/2]∙C6F6 + e− channel (open red circles). Solid black line is acquired by monitoring the total PE yield as a function of hv. The signal has been square rooted to account for the 2-photon nature of this feature. The red data points are acquired by taking the ratio of very low energy (indirect detachment) photoelectrons to high energy (direct detachment) photoelectrons.

I−C6F6 cluster: Ultrafast dynamics following charge-transfer The dynamics following excitation of the charge-transfer band just below the I[2P3/2]∙C6F6 + e− channel were probed by time-resolved photoelectron imaging9 (pump at 3.10 eV and probe at 1.55 eV) and the results have been reproduced here in Figure 13(a). Briefly, the peak at high eKE and at t = 0 was assigned to the formation of a transient nonvalence state of C6F6−. This non-valence state then evolves on a 30 fs timescale to the valence state of C6F6− (see Figure 5(b)). The oscillatory motion observed in the time-resolved PE spectra represents the buckling motion of C6F6− as it evolves from the planar geometry (associated with the non-valence state) to the buckled geometry (associated with the ground valence state – Figure 5). These dynamics could be conveniently tracked by integration over spectral ranges (highlighted in Figure 13(a)), as shown in Figure 13(b).


Page 23 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 13: (a) Time-resolved PE spectra of I−·(C6F6) excited at 3.10 and probed at 1.55 eV. t = 0 is indicated by a vertical dashed line. (b) Integrated photoelectron signal (open circles) over spectral windows indicated by the correspondingly coloured square brackets in (a). A fit to the oscillatory signal is included. This figure has been reproduced from reference 9.

Although the energy difference between the neutral and anion states in the planar geometry is small, our calculations show that this energy difference increases rapidly as the buckling angle increases. The energy difference can be attributed to a reduction in the energy 23 ACS Paragon Plus Environment


Page 24 of 37

of the anion as well as an increase in the energy of the neutral. From our C6F6− PE spectra, the difference between the anion and neutral at the anion valence state minimum is VDE = 1.60 eV (ignoring the presence of the iodine). At this point, some PE signal will still be observable based on a 1.55 eV probe and the width of the direct detachment peak (Figure 4). Experimentally, however, Figure 13(a) shows that the PE signal essentially disappears for eKE > 0.2 eV. This can be rationalised because the coherent motion naturally leads the molecule to buckle beyond its minimum energy geometry. While the anion state energy must increase, our calculations show that the corresponding neutral state energy is increasing more rapidly and therefore, the eKE measured by photodetachment with the probe decreases. Hence, the minimum of the pump-probe PE signal corresponds to a maximal buckling angle. Note that the above is based solely on energetic arguments and omits that the detachment cross-section along the buckling mode may also change. A specific (dominant) vibrational mode of e2u symmetry was assigned to the motion which has a calculated frequency (120 cm−1) that matches the measured beat frequency (121±2 cm−1). As part of the present study, we performed calculations in which we displaced the atoms along this vibrational mode vector (starting from the valence anionic minimum energy geometry). The results of this are shown in Figure 14 and indicate that the potential energy surface along this mode is mostly harmonic, consistent with the observed wavepacket motion in Figure 13. In addition to the oscillatory signal in Figure 13, dynamics can also be seen in the low eKE spectral range (eKE < 0.2 eV), which were not previously discussed. This low eKE peak has the appearance of an indirect detachment process and does not change in shape with pump-probe delay. This low eKE feature is not present in the single colour 3.10 eV PE spectrum and we therefore infer that the source of this low eKE peak must be a consequence of the interaction with the probe pulse. Its dynamics (integrated over 0.0 < eKE < 0.1 eV)


Page 25 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


show a step-like increase in the yield (see Figure 13(b)) and these steps are nearly π out-ofphase with the oscillatory dynamics observed in the 0.25 < eKE < 0.60 eV spectral window. Given that a probe photon is required for the production of the low eKE PE signal and that it appears to be an indirect process, we consider if excited states of C6F6− may be accessed by the probe photon. Figure 6 showed that resonances are available in C6 F6−. However, the energy of these resonances is based on C6F6− in its minimum energy geometry, while the dynamics access a wide range of buckle angles. We therefore considered the excited states of C6F6− at geometries away from the equilibrium of the valence state anion. Specifically, we calculated the excited states accessible at each geometry along the vibrational mode as shown in Figure 14. In the minimum energy geometry of the valence anion, the lowest energy bright transition involves a SOMO → LUMO excitation to the D2 excited state. The energy of this transition is > 2 eV and increases with greater buckle angle and is thus inaccessible the 1.55 eV probe. In contrast, the excited state corresponding to the excitation from HOMO → SOMO (D5 excited state in Figure 6), which is at > 3 eV at the minimum energy geometry, rapidly decreases in energy with a greater distortion along the vibrational buckling mode. While our calculations have not sampled the vibrational coordinate very accurately to predict the relevant energy profile of this excited state exactly, the trend is clear. As the molecule nears outer turning point, the transition energy to this state decreases. Hence, it seems reasonable to consider that near the outer turning point of the motion, the probe pulse may access this excited state. Note that at the extreme geometry, the excited state is bound (vertically) relative to the neutral and so the electron cannot detach directly. However, adiabatically, the excited state is unbound and we speculate that accessing this excited state eventually leads to autodetachment that produces the low eKE electrons observed in Figure 13.



Page 26 of 37

The above interpretation considered a simplified one-dimensional picture, lacking consideration of other vibrational modes that might be important, especially at the large geometric distortions in the extremes of the 1D cut along the buckling coordinate. The coupling to other modes is consistent with the dephasing of the wavepacket in Figure 13(a). It thus appears that, as other vibrational modes are populated along the buckling motion, the width of the transition increases and other electronic transitions become available along the entire coordinate, which would lead to the observed step-like signal. After a few vibrations along the buckling coordinate, it appears that much of the vibrational energy has redistributed into these unspecified modes, in agreement with the oscillatory motion in the 0.25 < eKE < 0.60 eV spectral window.

Figure 14: Potential energy curves along the vibrational coordinate of the wavepacket motion of C6F6−. The ground electronic states of C6F6− (D0 black) and C6F6 (S0 red) are shown as well as three excited states of C6F6−. The D2 excited state (green dashed) in the minimum energy geometry of C6F6− increases with buckle angle, while the D5 excited state (blue solid) decreases so that the probe may become resonant with a transition from D0 to this excited state. A third (D9) excited state is included (purple dashed), but probably plays no role in the measured dynamics.


Page 27 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 15: Time-resolved PE spectrum of I−(C6F6) taken at t = 0 with a 3.10 eV pump and 1.55 eV probe pulse. This spectrum is effectively the PE spectrum of the non-valence correlation-bound state of the (C6F6) −.

I∙C6F6−: Non-valence state photoelectron and photodetachment spectra We now return to the earliest dynamics following photoexcitation of I−(C6F6) at hv = 3.10 eV. A PE spectrum at t = 0 (equivalent to a vertical slice in Figure 13(a) at t = 0) is shown in Figure 15. This particular PE spectrum was acquired to achieve the best possible signal-to-noise ratio at t = 0. The PE spectrum shows an intense peak at eKE = 1.10 eV, which we have previously assigned to the non-valence state of C6F6−. There is also a clear peak at eKE = 0.16 eV, offset from the primary peak by the spin-orbit splitting of iodine. This peak may be assigned to a small contribution of detachment from the non-valence state (that is associated with the 2P3/2 state of I in I∙C6F6−) to the higher-lying 2P1/2 state of I in I∙C6F6−. There is a broad continuum of electrons between the two peaks due to rapid evolution of the non-valence to valence evolution within the finite time-resolution of the experiment9. Interestingly, there is a weak progression of peaks on the low eKE side of the peak assigned to direct detachment from the non-valence state. The first of these small peaks is 27 ACS Paragon Plus Environment


Page 28 of 37

shifted from the primary peak by 0.18 eV (~1,450 cm−1). The PE spectra of the CH3CN− dipole-bound anion measured by Bailey and coworkers exhibited similar vibrational structure45. These were assigned to vibrational bands that arise not only from differences in anion and neutral potential energy surfaces, but also from vibrational modes of the neutral that are strongly coupled to the non-valence state. In the case of the dipole-bound state in CH3CN−, these distinct peaks arise from the IR active modes because IR modes change the dipole-moment of the molecular core and therefore also the binding of the dipole-bound state. Bailey and coworkers were able to assign the individual peak offsets to particular IR-active vibrational modes of neutral CH3CN. We performed a similar analysis on the peaks observed in Figure 15 to ascertain whether they are likely to arise from a similar mechanism in which vibrational modes of C6F6 couple to the non-valence state. As the non-valence state of C6F6− is not dipole-bound but predominantly correlation-bound, we may naturally expect that the vibrational modes likely to be strongly coupled to the non-valence state are those that change the molecule’s polarizability. However, there is also a contribution from quadrupole-electron interaction in C6F6− and so changes in quadrupole moment may also influence non-valence state binding. Both of these physical properties are affected by Raman active modes. Furthermore, if C6 F6 distorts to produce a dipole-moment, then there would also be an attractive dipole-electron interaction. Therefore, all Raman and IR active vibrational modes could couple to the nonvalence state binding and might be responsible for the spectral features in Figure 15. The vibrational band (shifted by ~1,450 cm−1 from the direct detachment peak from the non-valence state) is close to both the 1,490 cm−1 Raman active a1g mode and the bright IR active 1,530 cm−1 e1u mode of planar (D6h) C6F646. Calculated vibrational modes of C6 F6 and of I∙C6F6 suggest, show that C6F6 clustering to iodine, red-shifts the a1g mode and the e1u modes by 30 cm−1 and 20 cm−1, respectively. Hence, both modes are very close to the


Page 29 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


measured peak shift. Unfortunately, it is not possible to establish whether either or both is observed in the experiment. There is also evidence for a second peak, shifted from the primary peak by twice the frequency of the first, 0.36 eV (~2,900 cm−1). This peak is probably an overtone of the a1g and/or e1u mode because C6F6 has no fundamental frequencies of this magnitude. Furthermore, we note that a peak at twice the frequency of the main vibration was also observed by Bailey et al. in the CH3CN− dipole-bound anion. Although they assigned them to IR modes of the neutral, it is also possible they are overtones of the main peaks (v3 and v6).

SUMMARY In summary, a new anion PE imaging instrument is described and used to probe the PE spectroscopy of (C6F6)n− (n = 1 – 5) over a large photon energy range. The PE images of the monomer showed that the PE emission is highly anisotropic for the D0 + hv → S0 + e− detachment channel, consistent with the electronic structure of the bent C2v geometry of C6F6−. We measure the VDE to be 1.60±0.07 eV and determined the upper limit for the ADE ≤ 0.70 eV, both consistent with previous measurements and calculations. We also observed a series of resonances manifest through the emission of photoelectrons with lower eKE than the direct detachment channel. The two lowest triplet states of C6F6 were observed and assigned to have a buckled nuclear geometry, similar to the anion ground state. The primary features that characterised the monomer spectra were also present in those of the clusters. Specifically, the D0 + hv → S0 + e− detachment channel was observed for n = 2 – 5 as well as the triplet states of the neutral and signatures of resonances. The increase in VDE as a function of the incremental increase in n was measured to be 200 meV



Page 30 of 37

in this range, in agreement with previous studies. Resonances were also apparent in the 2DPE spectra of clusters. The low energy PE emission seen upon the opening of the D0 + hv → T1/T2 + e− was argued to arise from the recapture of an outgoing photoelectron by a nonvalence anion state associated with the solvent molecules. The PE spectroscopy of I−(C6F6) was used to determine the rough shape of the absorption bands that appear just below the I[2P3/2]∙C6F6 + e− and I[2P1/2]∙C6F6 + e− direct detachment channels. Time-resolved PE spectra were used to probe the dynamics following excitation of the resonances just below the I[2P3/2]∙C6F6 + e− channel. This leads to the formation of a non-valence state that evolves into a valence state through a Jahn-Teller distortion along an out-of-plane buckling mode. The low energy indirect PE signal has been assigned to the excitation to an excited state of C6F6− at extreme buckling angles.

ACKNOWLEDGEMENTS This work was supported by the European Research Council through Starting Grant 306536.

REFERENCES [1]

Christophorou, L. G.; Datskos, P. G.; Faidas, H. Photodetachment in the Gaseous, Liquid, and Solid States of Matter. J. Chem. Phys. 1994, 101, 6728-6742.

[2]

Nyikos, L.; Ende, C. A. M.; Warman, J. M.; Hummel, A. High Mobility Excess Electrons in the Electron-Attaching Liquid Hexafluorobenzene. J. Phys. Chem. 1980, 84, 1154-1155.

[3]

Gahl, C.; Ishioka, K.; Zhong, Q.; Hotzel, A.; Wolf, M. Structure and Dynamics of Excited Electronic States at the Adsorbate/Metal Interface: C6F6/Cu(111). Faraday Discuss. 2000, 117, 191-202.


Page 31 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


[4]

Miller, T. M.; Doren, J. M. V.; Viggiano, A. A. Electron Attachment and Detachment: C6F6. Int. J. Mass Spectrom. 2004, 233, 67-73.

[5]

Hou, X.-J.; Huang, M.-B. Structure of the Hexafluorobenzene Anion. J. Mol. Struct.: THEOCHEM 2003, 638, 209-214.

[6]

Xie, Y.; Schaefer-III, H. F.; Cotton, F. A. The Radical Anions and the Electron Affinities of Perfluorinated Benzene, Naphthalene and Anthracene. Chem. Commun. 2002, 0, 102-103.

[7]

Gant, K. S.; Christophorou, L. G. Attachment of Slow Electrons to Hexafluorobenzene. J. Chem. Phys. 1976, 65, 2977-2981.

[8]

Voora, V. K.; Jordan, K. D. Nonvalence Correlation-Bound Anion States of Spherical Fullerenes. Nano Lett. 2014, 14, 4602-4606.

[9]

Rogers, J. P.; Anstöter, C. S.; Verlet, J. R. R. Ultrafast Dynamics of Low-Energy Electron Attachment via a Non-Valence Correlation-Bound State. Nat. Chem. 2018, 10, 341-346.

[10] Rogers, J. P.; Anstöter, C. S.; Verlet, J. R. R. Evidence of Electron Capture of an outgoing Photoelectron Wave by a Nonvalence State in (C6F6)n. J. Phys. Chem. Lett. 2018, 9, 2504-2509. [11] Dougherty, D. B.; Feng, M.; Petek, H.; Yates Jr, J. T.; Zhao, J. Band Formation in a Molecular Quantum Well via 2D Superatom Orbital Interactions. Phys. Rev. Lett. 2012, 109, 266802. [12] Voora, V. K.; Jordan, K. D. Nonvalence Correlation-Bound Anion States of Polycyclic Aromatic Hydrocarbons. J. Phys. Chem. Lett. 2015, 6, 3994-3997. [13] Field, D.; Jones, N. C.; Ziesel, J. P. Cold Electron Scattering in SF6 and C6F6: Bound and Virtual State Channels. Phys. Rev. A 2004, 69, 052716.



Page 32 of 37

[14] Suess, L.; Parthasarathy, R.; Dunning, F. B. Nondissociative Low-Energy Electron Attachment to SF6, C6F6, C10F8, and C-C7F14: Negative Ion Lifetimes. J. Chem. Phys. 2002, 117, 11222. [15] Eustis, S. N.; Wang, D.; Bowen, K. H.; Patwari, G. N. Photoelectron Spectroscopy of Hydrated Hexafluorobenzene Anions. J. Chem. Phys. 2007, 127, 114312. [16] Nakajima, A.; Taguwa, T.; Hoshino, K.; Sugioka, T.; Naganuma, T.; Oho, F.; Watanabe, K.; Nakao, K.; Konishi, Y.; Kishi, R.; Kaya, K. Photoelectron Spectroscopy of (C6F6)-n and (Au-C6F6)- Clusters. Chem. Phys. Lett. 1993, 214, 22-26. [17] Anstöter, C. S.; Bull, J. N.; Verlet, J. R. R. Ultrafast Dynamics of Temporary Anions Probed through the Prism of Photodetachment. Int. Rev. Phys. Chem. 2016, 35, 509538. [18] Bull, J. N.; West, C. W.; Verlet, J. R. R. Ultrafast Dynamics of Formation and Autodetachment of a Dipole-Bound State in an Open-Shell Pi-Stacked Dimer Anion. Chem. Sci. 2016, 7, 5352-5361. [19] Bull, J. N.; West, C. W.; Verlet, J. R. R. On the Formation of Anions: Frequency-, Angle-, and Time-Resolved Photoelectron Imaging of the Menadione Radical Anion. Chem. Sci. 2015, 6, 1578-1589. [20] Eppink, A. T. J. B.; Parker, D. H. Velocity Map Imaging of Ions and Electrons Using Electrostatic Lenses: Application in Photoelectron and Photofragment Ion Imaging of Molecular Oxygen. Rev. Sci. Inst. 1997, 68, 3477. [21] Even, U.; Jortner, J.; Noy, D.; Lavie, N.; Cossart-Magos, C. Cooling of Large Molecules below 1 K and He Clusters Formation. J. Chem. Phys. 2000, 112, 80688071. [22] Wiley, W. C.; McLaren, I. H. Time-of-Flight Mass Spectrometer with Improved Resolution. Rev. Sci. Inst. 1955, 26, 1150-1157.


Page 33 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


[23] Roberts, G. M.; Nixon, J. L.; Lecointre, J.; Wrede, E.; Verlet, J. R. R. Toward RealTime Charged-Particle Image Reconstruction Using Polar Onion-Peeling. Rev. Sci. Inst. 2009, 80, 053104. [24] Reid, K. L. Photoelectron Angular Distributions. Annu. Rev. Phys. Chem. 2003, 54, 397-424. [25] Ervin, K. M.; Anuslewicz, I.; Skurski, P.; Simons, J.; Lineberger, W. C. The Only Stable State of O2 Is the X 2Πg Ground State and It (Still!) Has an Adiabatic Electron Detachment Energy of 0.45 eV. J. Phys. Chem. A 2003, 107, 8521-8529. [26] Gordon, S. D. S.; Osterwalder, A. 3D-Printed Beam Splitter for Polar Neutral Molecules. Phys. Rev. Appl. 2017, 7, 044022. [27] Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, G. A.; Petersson, H., et al. Gaussian 09, Gaussian, Inc.: Wallingford, CT, 2009. [28] Yanai, T.; Tew, D. P.; Handy, N. C. A New Hybrid Exchange-Correlation Functional Using the Coulomb-Attenuating Method (CAM-B3LYP). Chem. Phys. Lett. 2004, 93, 51-57. [29] Dunning, T. H. Gaussian Basis Sets for Use in Correlated Molecular Calculations. I. The Atoms Boron through Neon and Hydrogen. J. Chem. Phys. 1989, 90, 1007-1023. [30] Runge, E.; Gross, E. K. U. Density-Functional Theory for Time-Dependent Systems. Phys. Rev. Lett. 1984, 52, 997-1000. [31] Casida, M. E. In Recent Advances in Density Functional Methods; Chong, D. P., Ed.; World Scientific: Singapore, 1995. [32] Petersilka, M.; Gossmann, U. J.; Gross, E. K. U. Excitation Energies from TimeDependent Density-Functional Theory. Phys. Rev. Lett. 1996, 76, 1212-1215.



Page 34 of 37

[33] Hirata, S.; Head-Gordon, M. Time-Dependent Density Functional Theory within the Tamm-Dancoff Approximation. Chem. Phys. Lett. 1999, 314, 291-299. [34] Hanwell, M. D.; Curtis, D. E.; Lonie, D. C.; Vandermeersch, T.; Zurek, E.; Hutchison, G. R. Avogadro: An Advanced Semantic Chemical Editor, Visualization, and Analysis Platform. J. Cheminf. 2012, 4, 17. [35] Anstöter, C. S.; Dean, C. R.; Verlet, J. R. R. Sensitivity of Photoelectron Angular Distributions to Molecular Conformations of Anions. J. Phys. Chem. Lett. 2017, 8, 2268-2273. [36] Wigner, E. P. On the Behavior of Cross Sections Near Thresholds. Phys. Rev. 1948, 73, 1002-1009. [37] Anstöter, C. S.; Dean, C. R.; Verlet, J. R. R. Chromophores of Chromophores: A Bottom-up Hückel Picture of the Excited States of Photoactive Proteins. Phys. Chem. Chem. Phys. 2017, 19, 29772-29779. [38] Stanley, L. H.; Anstöter, C. S.; Verlet, J. R. R. Resonances of the Anthracenyl Anion Probed by Frequency-Resolved Photoelectron Imaging of Collision-Induced Dissociated Anthracene Carboxylic Acid. Chem. Sci. 2017, 8, 3054-3061. [39] West, C. W.; Bull, J. N.; Hudson, A. S.; Cobb, S. L.; Verlet, J. R. R. Excited State Dynamics of the Isolated Green Fluorescent Protein Chromophore Anion Following UV Excitation. J. Phys. Chem. B 2015, 119, 3982-3987. [40] Nelson, D. J.; Gichuhi, W. K.; Miller, E. M.; Lehman, J. H.; Lineberger, W. C. Anion Photoelectron Spectroscopy of Deprotonated Ortho-, Meta-, and Paramethylphenol. J. Chem. Phys. 2017, 146, 074302. [41] Gunion, R. F.; Gilles, M. K.; Polak, M. L.; Lineberger, W. C. Ultraviolet Photoelectron Spectroscopy of the Phenide, Benzyl, and Phenoxide Anions. Int. J. Mass Spectrom. Ion Processes 1992, 117, 601-620.


Page 35 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


[42] Bochenkova, A. V.; Mooney, C. R. S.; Parkes, M. A.; Woodhouse, J. L.; Zhang, L.; Lewin, R.; Ward, J. M.; Hailes, H. C.; Andersen, L. H.; Fielding, H. H. Mechanism of Resonant Electron Emission from the Deprotonated GFP Chromophore and its Biomimetics. Chem. Sci. 2017, 8, 3154-3163. [43] West, C. W.; Bull, J. N.; Antonkov, E.; Verlet, J. R. R. Anion Resonances Probed by photoelectron Imaging of the Para-Benzoquinone Radical Anion. J. Phys. Chem. A 2014, 118, 11346-11354. [44] Mbaiwa, F.; Duzor, M. V.; Wei, J.; Mabbs, R. Direct and Indirect Detachment in the Iodide Pyrrole Cluster Anion: The Role of Dipole Bound and Neutral Cluster States. J. Phys. Chem. A 2010, 114, 1539-1547. [45] Bailey, C. G.; Dessent, C. E. H.; Johnson, M. A.; Bowen, K. H. Vibronic Effects in the Photon Energy-dependent Photoelectron Spectra of the CH3CN- Dipole Bound Anion. J. Chem. Phys. 1996, 104, 6976-6983. [46] Steele, D.; Whiffen, D. H. The Vibrational Frequencies of Hexafluorobenzene. Trans. Faraday Soc. 1959, 55, 369-376.



Page 36 of 37

Jan Verlet was born in Leuven, Belgium. He received his MSci and PhD from King’s College London in 2003 and was a post-doctoral fellow at the University of California at Berkeley. In 2006 he moved to Durham University as a lecturer and EPSRC Advanced Research Fellow and is he currently a Professor in physical chemistry. His research interests are focussed around ultrafast dynamics following photoexcitation of, photodetachment of, and electron attachment to molecules; and the development of methods to probe these dynamics in the gas-phase and at aqueous interfaces.


Page 37 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


The electronic structure and dynamics of anionic hexafluorobenzene and its clusters was studied using 2D photoelectron imaging. 82x44mm (300 x 300 DPI)

ACS Paragon Plus Environment

Photoelectron Spectroscopy of the Hexafluorobenzene Cluster Anions

Recommend Documents